<<

Chapter 4 Mobile Propagation (Large-scale Path Loss)

-50 +90

-60 +80 Multi-path Propagation -70 +70 Field -80 +60 Strength, RSSI, dBuV/m dBm -90 +50 -100 +40

-110 +30

-120 +20 0 3 6 9 12 15 18 21 24 27 30 33 Distance from Cell Site, km

measured signal strength

signal strength predicted by Okumura- Hata propagation model

Wireless Communication

Objectives

 To refresh understanding of basic concepts and tools  To discuss the basic philosophy of propagation prediction applicable to cellular systems  To identify and explore key propagation modes and their signal decay characteristics  To discuss the multi-path propagation environment, its effects, and a method of avoiding deep fades  To survey key available statistical propagation models and become familiar with their basic inputs, processes, and outputs  To understand application of statistical confidence levels to system propagation prediction  To review and gain familiarity with general measurement and propagation prediction tools available commercially

Chapter 4 –Mobile ( Large-scale path loss) 1 Dr. Sheng-Chou Lin

Page 1 Communication

Wave Propagation Basics: Frequency and Wavelength Wavelength is an important variable in RF propagation. /2  Wavelength determines the approximate required size of elements.

 Objects bigger than roughly a wavelength can reflect or block RF energy.

 RF can penetrate into an enclosure if it has holes roughly a wavelength in size, or larger.

Chapter 4 –Mobile radio propagation( Large-scale path loss) 2 Dr. Sheng-Chou Lin

Wireless Communication

Wave Propagation: Frequency and Wavelength

  Radio signals travel through empty space at the speed of light ( Cell C) •C = 186,000 miles/second (300,000,000 meters/second)  Frequency (F) is the number of waves per second (unit: Hertz) speed = C  Wavelength (length of one wave) is Examples: calculated: •(distance traveled in one second) /(waves AMPS cell site f = 870 MHz. in one second) 0.345 m = 13.6 inches

PCS-1900 site f = 1960 MHz. C/F 0.153 m = 6.0 inches

Chapter 4 –Mobile radio propagation( Large-scale path loss) 3 Dr. Sheng-Chou Lin

Page 2 Wireless Communication

Statistical Propagation Models

 Prediction of Signal Strength as a function of distance without regard to obstructions or features of a specific propagation path

-50 +90

-60 +80

-70 +70 Field -80 +60 Strength, RSSI, dBm dBuV/m -90 +50

-100 +40

-110 +30

-120 +20 0 3 6 9 12 15 18 21 24 27 30 33 Distance from Cell Site, km

measured signal signal strength predicted by Okumura- strength Hata propagation model

Chapter 4 –Mobile radio propagation( Large-scale path loss) 4 Dr. Sheng-Chou Lin

Wireless Communication

Radio Propagation

 Mobile radio channel •fundamental limitation on the performance of wireless communications.s •severely obstructed by building, mountain and foliage. •speed of motion •a statistical fashion  propagation characteristics •, diffraction and scattering •no direct line -of-sight path in urban areas •multipath  Basic propagation types •Propagation model: predict the average received signal strength •Large-scale fading: Shadowing fading •Small-scale fading: Multipath fading

Chapter 4 –Mobile radio propagation( Large-scale path loss) 5 Dr. Sheng-Chou Lin

Page 3 Wireless Communication

Propagation Model

 To focus on predicting the average received signal strength at a given distance from the •variability of the signal strength •is useful in estimating the radio coverage.  Large-scale propagation •computed by averaging over 540, 1m 10m, for1GHz 2GHz.  Small-scale fading •received signal strength fluctuate rapidly, as a mobile moves over very small distance. •Received signal is a sum of multi-path signals. • distribution •may vary by 30 40 dB •due to movement of propagation related elements in the vicinity of the receiver.

Chapter 4 –Mobile radio propagation( Large-scale path loss) 6 Dr. Sheng-Chou Lin

Wireless Communication

Deterministic Techniques Basic Propagation Modes  There are several very commonly- occurring modes of propagation, Free Space depending on the environment through which the RF propagates. Three are shown at right: •these are simplified, practically- calculable cases Reflection •real-world paths are often dominated by with partial cancellation one or a few such modes –these may be a good starting point for analyzing a real path –you can add appropriate corrections Knife-edge for specific additional factors you Diffraction identify •we’re going to look at the math of each one of these

Chapter 4 –Mobile radio propagation( Large-scale path loss) 7 Dr. Sheng-Chou Lin

Page 4 Wireless Communication

Free-Space Propagation

 Effective area (Aperture) Aeff = A d Aeff ratio of power delivered to the Pt antenna terminals to 1 the incident power Gt 2 Gr density •: Antenna efficiency P t S = Gt : power density 4 d2 •A : Physical area  Transmitter antenna P r = S A eff : Received power gain = Gt 2 4A eff  Receiver A eff = Gr  G = = G 4 2 r 2  Propagation distance = P t  P r = Gt G r  d 4 d2 4   Wave length = 

EIRP = PtGt (Effective isotropic radiated power)

Chapter 4 –Mobile radio propagation( Large-scale path loss) 8 Dr. Sheng-Chou Lin

Wireless Communication

Free-Space Propagation

 A clear, unobstructed Line-of-sight path between them •Satellite communication, Line-of-sight (Point-to-point)

antenna 1 G1 frequency f or wavelength  antenna 2 G2

P Pt r distance d

EIRP= PtGt = effective isotropic radiated power (compared to an ) : dBi ERP = EIRP-2.15dB = (compared to an half-wave ) : dBd Path Gain P 2  8 2 r   2  c   3 10  gain = ------= G G ------ = G G ------ = G G ------ P 1 24d  1 24df  1 2 3 6 t 4d 1 10 f 1 10  for d in km, f in MHz

Path Loss = 1 / (Pr/Pt) when antenna gains are included

loss(dB) = 32.44 + 20 log d + 20 log f –G 1 dB –G2 dB  Chapter 4 –Mobile radio propagation( Large-scale path loss) 9 Dr. Sheng-Chou Lin

Page 5 Wireless Communication

Free-Space Propagation

 The simplest propagation mode r •Imagine a transmitting antenna at the center of an empty sphere. Each little square of surface intercepts its share of the radiated energy •Path Loss, db (between two isotropic antennas) = 36.58 +20*Log10(FMHZ)+20Log10(DistMILES ) Free Space •Path Loss, db (between two dipole antennas) “Spreading”Loss = 32.26 +20*Log10(FMHZ)+20Log10(DistMILES ) energy intercepted •Notice the rate of signal decay: by the red square is • 6 db per octave of distance change, which is 20 db per decade of distance change proportional to 1/r2  When does free-space propagation apply? 1st •there is only one signal path (no reflections) d •the path is unobstructed (first Fresnel zone is not P penetrated by obstacles) A D First Fresnel Zone = B {Points P where AP + PB - AB < } Fresnel Zone radius d = 1/2 (D)^(1/2)

Chapter 4 –Mobile radio propagation( Large-scale path loss) 10 Dr. Sheng-Chou Lin

Wireless Communication Near and Far fields

D

/2 radiated 2 0 reactive 2D / radiated far field near field

 These distances are rough approximations!  Reactive near field has substantial reactive components which die out  Radiated near field angular dependence is a function of distance from the antenna (i.e., things are still changing rapidly)  Radiated far field angular dependence is independent of distance  Moral: Stay in the far field!

Chapter 4 –Mobile radio propagation( Large-scale path loss) 11 Dr. Sheng-Chou Lin

Page 6 Wireless Communication

An Example

 An antenna with maximum dimension (D) of 1m, operating frequency (f) = 900 MHz. •= c/f = 3108/900 106 = 0.33 2 2 •Far-field distance = df = 2D / = 2 (1) /0.33 = 6m

 TX power, Pt = 50W, fc = 900MHz, Gt = 1 = Gr 3 •Pt (dBm) = 10log(50 10 mW) = 47 dBm = 10lon(50) = 17 dBW

•Gt = 1 = Gr = 0dB

•Loss (100m)= 32.44 +20log(dkm)+20log(fMHz) = 32.44 + 20log (0.1)+20log(900) =71.525 dB

–Pr (100m)= 47 +0 –71.525 +0 = -24.5 dBm

•Loss (10km)= 32.44 +20log(dkm)+20log(fMHz) = 32.44 + 20log (10)+20log(900) =71.525 dB + 40 = 111.525 dB

–Pr (100m)= 47 +0 –111.525 +0 = -64.5 dBm

Chapter 4 –Mobile radio propagation( Large-scale path loss) 12 Dr. Sheng-Chou Lin

Wireless Communication

A Tedious Tale of One Radio Link

Transmitter Let’s track the power flow from 20 TX output transmitter to receiver in the x 0.50 line efficiency radio link we saw back in lesson Trans. = 10 watts to antenna x 20 antenna gain 2. We’re going to use real values Line = 200 watts ERP that commonly occur in typical x 0.000,000,000,000,000,1585links.path Antenna = 0.000,000,000,000,031,7 watts if intercepted by dipole antenna

x 20 antenna gain = 0.000,000,000,000,634 watts into line

Antenna x 0.50 line efficiency = 0.000,000,000,000,317 watts to receiver Trans. Line Did you enjoy that arithmetic? Let’s go back and do it again, a better and less painful way.

Receiver

Chapter 4 –Mobile radio propagation( Large-scale path loss) 13 Dr. Sheng-Chou Lin

Page 7 Wireless Communication

Decibels - A Helpful Convention

x 1000  normally refer to power ratios -- in other words, the numbers we represent in dB usually .001 w 1 are a ratio of two powers. Examples: 0 dBm 30 dBm •A certain amplifies its input by a factor of +30 dB 1,000. (Pout/Pin = 1,000,000). That amplifier has 30 dB gain. x 0.10 10 w •A certain has an efficiency of only 10 100 w percent. (P /P = 0.1) The transmission line has a +50 dBm +40 dBm out in -10 dB loss of -10 dB.  Often decibels are used to express an absolute • dB are comfortable-size numbers • rather than multiply and divide RF number of watts, milliwatts, kilowatts, etc. When power ratios, in dB we can just add used this way, we always append a letter (W, m, or & subtract • Given a number, convert to dB: K) after “db”to show the unit we’re using. For

db = 10 x Log10 (N) example, • Given dB, convert to a number: •20 dBK = 50 dBW = 80 dBm= 100,000 watts N = 10^(db/10) •0 dBm = 1 milliwatt

Chapter 4 –Mobile radio propagation( Large-scale path loss) 14 Dr. Sheng-Chou Lin

Wireless Communication A Much Less Tedious Tale of that same Radio Link

Transmitter +43 dBm TX output Let’s track the power flow again, using decibels. Trans. -3 dB line efficiency Line = +40 dBm to antenna +13 dB antenna gain = +53 dBm ERP Antenna -158 dB path attenuation = -105 dBm if intercepted by dipole antenna (+4.32dB for EIRP)

+13 dB antenna gain = -92 dBm into line Antenna -3 dB line efficiency = -95 dBm to receiver Trans. Line  Wasn’t that better?! Let’s look at how dB work.

Receiver

Chapter 4 –Mobile radio propagation( Large-scale path loss) 15 Dr. Sheng-Chou Lin

Page 8 Wireless Communication

Introduction The Function of an Antenna

Antenna 1 Antenna 2

Transmission Electromagnetic Transmission Line Field Line RF RF Power Power current current Available

An antenna is a passive device (an arrangement of electrical conductors) which converts RF power into electromagnetic fields, or intercepts electromagnetic fields and converts them into RF power.  RF power causes current to flow in the antenna.  The current causes an electromagnetic field to radiate through space.  The electromagnetic field induces small currents in any other conductors it passes. These currents are small, exact replicas of the original current in the original antenna.

Chapter 4 –Mobile radio propagation( Large-scale path loss) 16 Dr. Sheng-Chou Lin

Wireless Communication

Antenna Antenna 1 Vertically Antenna 2 Polarized Horizontally Polarized Transmission Electromagnetic Transmission Line Field Line RF RF Power Power current Available almost no current

Antenna 1

 The electromagnetic field is oriented by the direction of current flow in the radiating antenna.  To intercept significant energy, a receiving antenna should be oriented parallel to the transmitting antenna. A receiving antenna oriented at right angles to the transmitting antenna will have very little current induced in it. This is referred to as cross- polarization. Typical cross-polarization loss is 20 dB.  Vertical polarization is the norm in mobile telephony.

Chapter 4 –Mobile radio propagation( Large-scale path loss) 17 Dr. Sheng-Chou Lin

Page 9 Wireless Communication

Reference Antennas and Effective Radiated Power

Effective Radiated Power is always expressed in relation to the radiation produced by a reference antenna. Isotropic  The flashlight example used a plain light bulb as Antenna a reference - producing the same light in all directions.  The radio equivalent of a plain light bulb is called an isotropic radiator. It radiates the same in all directions. Unfortunately, it virtually impossible to build such an antenna. •Radiation compared to an isotropic radiator is called Dipole Antenna EIRP, Effective Isotropic Radiated Power. Null  The simplest, most common, physically constructible reference antenna is a dipole. Main •Radiation compared to a dipole is called ERP, Effective Lobe Radiated Power. Null

Chapter 4 –Mobile radio propagation( Large-scale path loss) 18 Dr. Sheng-Chou Lin

Wireless Communication

Reference Antennas, ERP and EIRP Dipole Null

 ERP is by comparison to a Dipole •This is the tradition in cellular, land mobile, HF Null communications, and FM/TV Isotropic  EIRP is by comparison to an Isotropic Radiator •This is the tradition in PCS at 1900 MHz., microwave, satellite communications, and  ERP values can be converted to EIRP and vice versa. •For a given amount of power input, a dipole produces 2.16 db more radiation than an isotropic radiator, due to the dipole slight directionality. A third antenna compared against both dipole and isotropic will have a bigger EIRP (vs. isotropic) than ERP (vs dipole). The difference is 2.16 db, a power ratio of 1.64. Therefore,

ERP = EIRP - 2.16 dB and ERP = EIRP / 1.64 EIRP = ERP + 2.16 dB and EIRP = ERP x 1.64

Chapter 4 –Mobile radio propagation( Large-scale path loss) 19 Dr. Sheng-Chou Lin

Page 10 Wireless Communication

Radiation Patterns Key Features and Terminology

Radiation patterns of antennas are usually Typical Example plotted in polar form Horizontal Plane Pattern  The Horizontal Plane Pattern shows the Notice -3 dB points radiation as a function of azimuth 0 (N) (i.e.,direction N-E-S-W) 0 10 db  The Vertical Plane Pattern shows the -10 points radiation as a function of elevation -20 Main (i.e., up, down, horizontal) -30 db Lobe  Antennas are often compared by 270 90 noting specific features on their (W) nulls or a Minor (E) patterns: minima Lobe •-3 db (“HPBW”), -6 db, -10 db points Front-to-back Ratio •front-to-back ratio •angles of nulls, minor lobes, etc. 180 (S)

Chapter 4 –Mobile radio propagation( Large-scale path loss) 20 Dr. Sheng-Chou Lin

Wireless Communication

Two Basic Methods of Obtaining Gain

Quasi-Optical Techniques (reflection, focusing) •Reflectors can be used to concentrate radiation –technique works best at microwave frequencies, where reflectors are small •examples: –corner reflector used at cellular or higher frequencies –parabolic reflector used at microwave frequencies –grid or single pipe reflector for cellular Array Techniques (discrete elements) •power is fed or coupled to multiple antenna elements; each element radiates In Phase •elements?radiations in phase in some directions •in other directions, different distances to distant observer introduce different phase delay for each Out of element, and create pattern lobes and nulls Phase

Chapter 4 –Mobile radio propagation( Large-scale path loss) 21 Dr. Sheng-Chou Lin

Page 11 Wireless Communication

Real-World Path Loss

Free space is, in general, NOT the real world. We must deal with:

•reflections over flat or curved Earth •reflections from smooth •scattering from rough surfaces •diffraction around/over obstacles •absorption by vegetation and other lossy media, including buildings and walls •multipath fading •approximately fourth power propagation loss

Chapter 4 –Mobile radio propagation( Large-scale path loss) 22 Dr. Sheng-Chou Lin

Wireless Communication

Reflection

 A propagating wave impinges upon an object with very large dimensions ( >> )  Reflections occur from surface of the earth and from building and wells  Flat surface

Path Loss (dB) = 40Log(d) - [10Log(Gt )+10Log (Gr ) +20Log (hb ) +20Log (hm )]

2 2 hb hm P P G G r = t t  r 4 TX ERPDBM d

 = r1 + r2 - r = phase difference in two paths h - h r b m h b r1

 r2 d

Chapter 4 –Mobile radio propagation( Large-scale path loss) 23 Dr. Sheng-Chou Lin

Page 12 Wireless Communication Reflection with Partial Cancellation

 Assumptions: •the cell is a mile away or more •the cell is not over a few hundred feet Direct ray higher than the car •there are no other obstructions Reflected  If these assumptions are true, then: Ray •The point of reflection will be very close to the car -- at most, a few hundred feet away. •the difference in path lengths is influenced Point of most strongly by the car antenna height reflection above ground or by slight ground height variations  The reflected ray tends to cancel the This reflection is at “grazing incidence”. The reflection is virtually 100% efficient, direct ray, dramatically reducing the and the phase of the reflected signal flips received signal level 180 degrees.

Chapter 4 –Mobile radio propagation( Large-scale path loss) 24 Dr. Sheng-Chou Lin

Wireless Communication

Reflection with Partial Cancellation  Analysis: •physics of the reflection cancellation TX ERPDBM predicts signal decay approx. 40 db per decade of distance –twice as rapid as in free-space! HTFT HT •observed values in real systems range FT from 30 to 40 db/decade Received Signal Level, dBm =

TX ERPDBM - 172 DMILES - 34 x Log10 (DMILES ) + 20 x Log10 (Base Ant. HtFEET) + 10 x Log10 (Mobile Ant. HtFEET) Comparison of Free-Space and Reflection Propagation Modes Assumptions: Flat earth, TX ERP = 50 dBm, @ 870 MHz. Base Ht = 200 ft, Mobile Ht = 5 ft.

DistanceMILES 1 2 4 6 8 10 15 20

FS usingFree-SpaceDBM -45.3 -51.4 -45.3 -57.4 -63.4 -65.4 -68.9 -71.4

FS using ReflectionDBM -69.0 -79.2 -89.5 -95.4 -99.7 -103.0 -109.0 -113.2

Chapter 4 –Mobile radio propagation( Large-scale path loss) 25 Dr. Sheng-Chou Lin

Page 13 Wireless Communication

Observation on Signal Decay Rates

Signal Level vs. Distance We’ve seen how the signal decays 0 with distance in two simplified modes of propagation: -10  Free-Space •20 dB per decade of distance -20 •6 db per octave of distance  Reflection Cancellation -30 •40 dB per decade of distance •12 db per octave of distance -40 1 2 3.16 5 6 7 8 10  Real-life cellular propagation decay Distance, Miles rates are typically somewhere One Decade between 30 and 40 dB per decade One Octave of distance (10x) of distance (2x) of distance

Chapter 4 –Mobile radio propagation( Large-scale path loss) 26 Dr. Sheng-Chou Lin

Wireless Communication

Diffraction

 Diffraction allows radio signals to propagate around the curved surface of the earth, beyond the horizon, and to propagate behind obstructions. •The diffraction field still exists and often has sufficient strength to produce a useful signal, as a receiver moves deeper into the obstructed (shadowed) region. •Caused by the propagation of secondary wavelets into a shadowed region. •Sum of the electric field components of all the secondary wavelets in the space around the obstacle.  Excess path length (  ): the difference between the direct path and the diffracted path. •A function of height and position of the obstruction, as well as the transmitter and receiver location.  Fresnel zones: successive receiver where  = n/2 •provide constructive and destructive interference to the total received signal. •Obstruction does not block the volume within the first Fresnel zone.

Chapter 4 –Mobile radio propagation( Large-scale path loss) 27 Dr. Sheng-Chou Lin

Page 14 Wireless Communication

Diffraction parameter

2  Excess path length (difference between h ( d1 + d2 ) = direct path and diffracted path) 2 d1 d2 2 2 2 h ( d1 + d2 ) •The corresponding phase difference = =   2 d1 d2 d + d •= +    h ( 1 2 ) d1 d2 2 d1 d2 •Fresnel-Kirchoff diffraction Parameter =  ( d + d ) 1 2

  = 2 2  h  Phase difference bet. LOS and diff. Path is a function of height and position of the obstruction, as well as TX and RX d1 d2

Chapter 4 –Mobile radio propagation( Large-scale path loss) 28 Dr. Sheng-Chou Lin

Wireless Communication Fresnel Zones R

) 0 d1 d2 B d (

s 4 s o l 8 n o i t c

a 12 r

Rx f f

Tx i obstacle D 16 nd d Fresnel zones are the family of ellipsoids 1 2 Rn = certain values of phase of the rays which (d1+d2) h 0 which are the loci of circles that indicate as = n  h = 0 is the direct ray pass through them or = n /2  Generally want antenna heights high enough so all obstacles are below first Fresnel zone (n = 1)  If tip of obstacle is at center of Fresnel zone (LOS ray), then loss is 6 dB greater than free-space path loss

d1d2 R =  R is 1st Fresnel Zone radius, d1,d2 in km, and f in GHz (d1+d2) f

Chapter 4 –Mobile radio propagation( Large-scale path loss) 29 Dr. Sheng-Chou Lin

Page 15 Wireless Communication Knife-Edge Diffraction

 Radio signals to propagate between Transmitter H and Receiver is obstructed by a surface that has sharp irregularities (edges) such as hill or mountain.

R1 R2  Sometimes a single well-defined obstruction blocks the path. This case is fairly easy to analyze 2 1 1 and can be used as a manual tool to estimate the = H  effects of individual obstructions.  R1 R2   First calculate the parameter from the geometry of the path Ed /Eo = F(), Gd,dB= 20logF()   Next consult the table to obtain the obstruction 0 loss in db -5  Add this loss to the otherwise-determined path -10 atten loss to obtain the total path loss. dB -15 -20  Other losses such as reflection cancellation still -25 apply, but computed independently for the path sections before and after the obstruction. -5 -4 -3 -2 -1 0 1 2 3  Chapter 4 –Mobile radio propagation( Large-scale path loss) 30 Dr. Sheng-Chou Lin

Wireless Communication An Example of Diffraction

 = 1/3 m, d1 = 1km, d2 =1km,and(a)h=25m(b)h =0(c)h= -25m.

2 d1 d2 (a) h = 25m: = 2.74, Loss = 22 dB from =  Figure 4.14. Approximation = 21.7 dB, = ( d + d )  1 2 0.625m, = 1/3, n = 3.75  the tip of the 2 obstruction completely blocks the first h ( d1 + d2 ) = three Fresnel zones. 2 d1 d2 (b) h = 0m: = 0, Loss = 6 dB from Figure 4.14. Approximation = 6 dB, = 0m  the = n or = n /2 for Fresnel tip of the obstruction lies in the middle of Zones the first Fresnel zone.

(c) h = -25m: = -2.74, Loss = 1dB from h Figure 4.14. Approximation = 0dB, = 0.625m, = 1/3, n = 3.75  the tip of the obstruction completely blocks the first three Fresnel zones. However, the diffraction losses are negligible, since the d d 1 2 obstruction is below the LOS.

Chapter 4 –Mobile radio propagation( Large-scale path loss) 31 Dr. Sheng-Chou Lin

Page 16 Wireless Communication

Scattering

 Why consider scattering •Actual received signal what predicted by reflection and diffraction •Rough surface  reflected energy is spread out (diffused) •Flat surface with dimensions . •number of obstacles per unit volume is large. •Rough surfaces, small objects  irregularities scattering •ex. Foliage, trees, , street signs, lamp post.

 Rayleigh criterion

•Rough surface: h > hc, hc = / 8sinI

•reflection coefficient = flat coefficient s

–rough = s 

–s : scattering loss

Chapter 4 –Mobile radio propagation( Large-scale path loss) 32 Dr. Sheng-Chou Lin

Wireless Communication

Propagation Model

 General types •Outdoor •Indoor : conditions are much more variable.  Most of these models are based on a systematic interpretation of measurement data obtained in the service area.  Parameters used in propagation model •Frequency •Antenna heights •Environments : Large city, medium city, suburban, Rural (Open) Area.  Common models • : 20km > Range >1km •Walfisch and Bertoni Model :Range < 5km •Indoor propagation models : include scattering, reflection, diffraction –conditions are much more variable

Chapter 4 –Mobile radio propagation( Large-scale path loss) 33 Dr. Sheng-Chou Lin

Page 17 Wireless Communication

Statistical Propagation Models

 Based on statistical analysis of large amounts of measurement data  Predict signal strength as a function of distance and various parameters  Useful for early network dimensioning, number of cells, etc.  “Blind”to specific physics of any particular path -- based on statistics only  Easy to implement as a spreadsheet on PC or even on hand- held programmable calculator  Very low confidence level if applied as spot prediction method, but very good confidence level for system-wide generalizations

Chapter 4 –Mobile radio propagation( Large-scale path loss) 34 Dr. Sheng-Chou Lin

Wireless Communication

Statistical Propagation Models

 Prediction of Signal Strength as a function of distance without regard to obstructions or features of a specific propagation path

-50 +90

-60 +80

-70 +70 Field -80 +60 Strength, RSSI, dBm dBuV/m -90 +50

-100 +40

-110 +30

-120 +20 0 3 6 9 12 15 18 21 24 27 30 33 Distance from Cell Site, km measured signal strength signal predicted by Okumura- strength Hata propagation model Chapter 4 –Mobile radio propagation( Large-scale path loss) 35 Dr. Sheng-Chou Lin

Page 18 Wireless Communication Statistical Propagation Models: Commonly-required Inputs

 Frequency  Distance from transmitter to receiver  Effective Base Station Height  Average Terrain Elevation  Arbitrary loss allowances based on rules-of-thumb for type of area (Urban, Suburban, Rural, etc.)  Arbitrary loss allowance for penetration of buildings/vehicles  Assumptions of statistical distribution of variation of field strength values

Chapter 4 –Mobile radio propagation( Large-scale path loss) 36 Dr. Sheng-Chou Lin

Wireless Communication

Okumura Model

L50 (dB) = LF +Amu (f,d) –G(ht) –G(hr) –GAREA

• Widely used model for signal prediction in urban areas • is based on measured data and does not provide any analytical explanation

Where: L50 = The 50% (median) value of propagation path loss LF = The free space propagation loss Amu (f,d) = median attenuation relative to free space (see Fig. 3.23) G(ht) = Base station antenna height gain factor (30m ~1000m) G(hr) = mobile antenna height gain factor GAREA = Gain due to the type of environment (see Fig. 3. 24)

f : 150MHz ~ 1920MHz (up to 3000MHz), d: 1km ~ 100km

G(ht) = 20log ( ht /200 ), G(hr) = 10log ( hr /3 ), hr 3m G(hr) = 20log ( hr /3 ), 10m hr 3m

Chapter 4 –Mobile radio propagation( Large-scale path loss) 37 Dr. Sheng-Chou Lin

Page 19 Wireless Communication An Example using

 D= 50 km, ht = 100m, hr = 10m, in an urban environment. EIRP = 1kW, f = 900 MHz, unit gain receiving antenna.

• LF = 125.5dB

• Amu(900MHz, 50 km)) = 43 dB

• GAREA = 9dB

• G(ht) = -6dB

• G(hr) = 10.46 dB

• L50 = 155.04 dB

• Pr(d) = 60-155.04 + 0 = -95.04 dBm

Chapter 4 –Mobile radio propagation( Large-scale path loss) 38 Dr. Sheng-Chou Lin

Wireless Communication

Hata Model

L50 (Urban) (dB) = 69.55 + 26.16 log (F) –13.82 log(Hb) + (44. 9 –6.55 log(Hb) )*log (D) –a

Where: A = Path loss F = Frequency in mHz (150M-1500 MHz) D = Distance between base station and terminal in km (1km ~20km) H = Effective height of base station antenna in m (30m ~200m) a = Environment correction factor for mobile antenna height (1m~10m)

a = (1.1 log (F) - 0.7) Hm - (1.56 log (F)- 0.8 ) dB = Small~medium sized city (urban)

8.29 (log ( 1.54 Hm)) 2 - 1.1 dB for F 300 MHz = Large city (Dense 2 3.2 log (F) (log (11.75 Hm)) - 4.97 dB for F 300 MHz Urban)

2 • L50 (Urban) - 2(log(F/28)) - 5.4 = Suburban 2 • L50 (Urban) - 4.78(log(F)) - 18.33 (log(F)) - 40.98 = Rural (open)

• L90 = L50 + 10.32 dB : 90% QOS, L50 is the median value of propagation loss

Chapter 4 –Mobile radio propagation( Large-scale path loss) 39 Dr. Sheng-Chou Lin

Page 20 Wireless Communication

COST-231 Hata Model

A (dB) = 46.3 + 33.9log (F) –13.82 log(Hb) + (44. 9 –6.55 log(Hb) )*log (D) –a + c

Where: A = Path loss F = Frequency in MHz (1500M-2000 MHz) D = Distance between base station and terminal in km (1km ~20km) H = Effective height of base station antenna in m (30m ~200m) a = Environment correction factor for mobile antenna height c = Environment correction factor

C = 0 dB = Small~medium sized city (urban), Suburban

3 dB = Dense Urban (metropolitan center)

A is defined in the Hata Model

Chapter 4 –Mobile radio propagation( Large-scale path loss) 40 Dr. Sheng-Chou Lin

Wireless Communication

Statistical Propagation Models Okumura-Hata Model

A (dB) = 69.55 + 26.16 log (F) –13.82 log(H) + (44. 9 –6.55 log(H) )*log (D) + C

Where: A = Path loss F = Frequency in MHz (800-900 MHz) D = Distance between base station and terminal in km H = Effective height of base station antenna in m C = Environment correction factor

C = 0 dB = Dense Urban - 5 dB = Urban - 10 dB = Suburban - 17 dB = Rural

Chapter 4 –Mobile radio propagation( Large-scale path loss) 41 Dr. Sheng-Chou Lin

Page 21 Wireless Communication

Statistical Propagation Models COST-231 HATA Model

A (dB) = 46.3 + 33.9*logF –13.82*logH + (44.9 –6.55*logH)*log D + C

Where: A = Path loss F = Frequency in mHz (between 1700 and 2000 mHz) D = Distance between base station and terminal in km H = Effective height of base station antenna in m C = Environment correction factor

C = - 2 dB = for dense urban environment: high buildings, medium and wide streets

- 5 dB = for medium urban environment: modern cities with small parks

- 8 dB = for dense suburban environment, high residential buildings. wide streets

- 10 dB = for medium suburban environment. industrial area and small homes

- 26 dB = for rural with dense forests and quasi no hills

Chapter 4 –Mobile radio propagation( Large-scale path loss) 42 Dr. Sheng-Chou Lin

Wireless Communication

Statistical Propagation Models Walfisch-Ikegami Model

 Useful only in dense urban environments, but often superior to other methods in this environment  Based on “urban canyon”assumption • a “carpet”of buildings divided into blocks by street canyons

-20 • Uses diffraction and reflection mechanics and Signal Level -30 dbm statistics for prediction Legend -40 dbm -50 dbm -60 dbm • Input variables relate mainly to the geometry of the -70 dbm buildings and streets -80 dbm -90 dbm -100 dbm  Useful for two distinct situations: -110 dbm Area View -120 dbm • macro-cell - antennas above building rooftops • micro-cell - antennas lower than most buildings  Available in both 2-dimensional and 3-dimensional versions

Macrocell Microcell

Chapter 4 –Mobile radio propagation( Large-scale path loss) 43 Dr. Sheng-Chou Lin

Page 22 Wireless Communication

Statistical Techniques Practical Application of Distribution Statistics Percentage of Locations where Observed  Technique: RSSI exceeds Predicted RSSI •use a model to predict RSSI 10% of locations exceed this RSSI •compare measurements with model 50% – obtain median signal strength RSSI, 90% – obtain standard deviation dBm – now apply correction factor to obtain field strength required for desired probability of service  Applications: Given Distance •a desired signal level Normal Occurrences •the standard deviation of signal strength Distribution measurements •a desired percentage of locations which Median RSSI must receive that signal level Signal •We can compute a “cushion”in dB Strength , which will give us that % coverage dB

Chapter 4 –Mobile radio propagation( Large-scale path loss) 44 Dr. Sheng-Chou Lin

Wireless Communication

Propagation Loss

 Comparison among models •Free space •Hata Model (Okumura + COST 231) • Walfisch : considered by ITU-R in IMT-2000 standard. Cellular system, f = 850MHz PCS system, f =1900 MHz

Propagation Distance Propagation Distance 80.00 80.00 Dense Urban (Hata) Dense Urban (Hata) Urban (Hata) Urban (Hata) Suburban Suburban Rural 100.00 100.00 Rural Free Space Free Space Dense Urban (Walfish) Dense Urban (Walfish) Urban (Walfish) Urban (Walfish) 120.00 120.00 ) ) B B d d ( (

s s

s 140.00

s 140.00 o o L L

160.00 160.00

180.00 180.00

200.00 200.00 0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20 Distance (km) Distance (km)

Chapter 4 –Mobile radio propagation( Large-scale path loss) 45 Dr. Sheng-Chou Lin

Page 23 Wireless Communication Statistical Techniques Example of Application of Distribution Statistics  Suppose you want to design a Standard Gaussian ( m = 0, =1) cell site to deliver at least -95 Cumulative Normal Distribution dBm to at least 90% of the 100% locations in an area 90%  Measurements you’ve made 80% have a 10 dB. standard deviation 70% above and below the average 60% signal strength 50%  On the chart: 40% •to serve 90% of possible 30% locations, we must deliver an

20% average signal strength 1.29 standard deviations stronger than 10% -95 dBm, = 10 0% •-95+(1.29x10)= - 82 dbm -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 Standard Deviations from •Design for an average signal Median (Average) Signal Strength strength of - 82 dbm!

Chapter 4 –Mobile radio propagation( Large-scale path loss) 46 Dr. Sheng-Chou Lin

Wireless Communication

Statistical Techniques Normal Distribution Graph & Table for Convenient Reference

Cumulative Normal Distribution Standard Cumulative Deviation Probability 100% -3.09 0.1% 90% -2.32 1% 80% -1.65 5% 70% -1.28 10% 60% -0.84 20% -0.52 30% 50% 2.35 99% 40% 0 50% 30% 0.52 70% 20% 0.84 80% 1.28 90% 10% 1.65 95% 0% 2.35 99% -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.09 99.9% Standard Deviations from Mean Signal Strength

Chapter 4 –Mobile radio propagation( Large-scale path loss) 47 Dr. Sheng-Chou Lin

Page 24 Wireless Communication

Log-normal Shadowing Fading

 Long-term variation are due to propagation through obstructions, ets. And if the number of obstructions is large

 The loss in dB is respected as a Gaussian distribution with a mR(dB) and variance (dB)

•mR is the median loss of the path  We choose a specific coverage criteria such as 95%, 90%, 85% . To find 90% Loss •Pr [ Loss m + Loss ()] = 90% R Normal Occurrences Distribution  In all these cases, itself is a function of the environment •Large, medium city, suburban : 8dB Median RSSI Signal •Rural area 4dB Strength , –for = 9 dB.and 90% coverage, Loss ()= dB 10.32 dB

Chapter 4 –Mobile radio propagation( Large-scale path loss) 48 Dr. Sheng-Chou Lin

Wireless Communication

Shadowing Fading statistics

 Long-term variation is modeled by a log-normal distribution

- (j) d - d Ep = Eo e  Ep = Eo e

•at the receiver, the input signal will be given by Eo Ep

- i ri - iri d Er = Ei e  Er = Ei e

101og(Y) = 10log(e)X

•if number of obstructions is large, - I ri is Gaussianly distributed for any  and r Normal I i Distribution •y = ex: , y is lognormally distributed if x is Gaussianly distributed. , dB Chapter 4 –Mobile radio propagation( Large-scale path loss) 49 Dr. Sheng-Chou Lin

Page 25 Wireless Communication

Percentage of Coverage

 The percentage of area with a received signal , I.e. Pr [ Pr (R) ] •: desired received signal threshold • distance from the transmitter –received signal at D = R exceeds the threshold  –see Fig. 3.18 for different n and   Ex: shadowing deviation = 8dB, 75% boundary coverage (QOS) •Loss exponent factor n = 4  area coverage 94% •Loss exponent factor n = 2  area coverage 91%

Chapter 4 –Mobile radio propagation( Large-scale path loss) 50 Dr. Sheng-Chou Lin

Wireless Communication

Statistical Propagation Models Typical Results

Example of Model Results: Typical Cell Range Predictions for Various Environments

Tower Height EIRP Range F = 1900 mHz (meters) (watts) (km) Dense Urban 30 200 1.05 Urban 30 200 2.35 Suburban 30 200 4.03 Rural 50 200 10.3

Chapter 4 –Mobile radio propagation( Large-scale path loss) 51 Dr. Sheng-Chou Lin

Page 26 Wireless Communication

Building Penetration Losses

 Usual technique for path loss into a building: get median signal level in streets by some “normal”method add building penetration losses  Loss 1/h in general  Small scale variation is Rayleigh  Large scale variation is log-normal  Loss 1/f  Each additional floor is about 2 dB difference in loss  For primarily scattering paths, standard deviation is about 4 dB  For paths with at least partial LOS, standard deviation is about 6 to 9 dB  Windows of many new buildings have a thin layer of metal sputtered on the window glass; this increases attenuation

D. Molkdar, “Review on radio propagation into and within buildings,”IEE Proc-H, Vol. 138, No. 1, Feb 1991, pp 61- 73.

Chapter 4 –Mobile radio propagation( Large-scale path loss) 52 Dr. Sheng-Chou Lin

Wireless Communication

Propagation Inside Buildings

 Indoor environment differs from mobile environment in • interference environment (usually higher, due to equipment) • fading rate (usually slower, due to reduced speeds)  Limitations due to bandwidth • narrowband (e.g. TDMA) systems coverage limited by multipath and shadow fading • wideband systems experience ISI due to delay spread (less frequency diversity gain)  Power as a function of distance varies over a range: • P 1/d2 in a near-free-space environment (hallway) • P 1/d6 in a high-clutter environment (room full of cubes)  Loss: floors with structural metal > brick > plaster wall  Office fading is usually more continuous /smaller dynamic range than mobile fading  Stairwells and elevator shafts can act as and aid floor-to-floor propagation  Presence of an LOS path reduces RMS delay spread

Chapter 4 –Mobile radio propagation( Large-scale path loss) 53 Dr. Sheng-Chou Lin

Page 27 Wireless Communication Propagation Inside Buildings - Prediction

Tx

actual

direct diffraction Rx number of oors dif fracted path direct path  Combination of ray-tracing and diffraction can be very accurate at predicting inside propagation  Direct rays (through floors): each floor increases loss  Diffraction (windows and outside): large loss initially, but more floors do not add much loss  900 MHz band losses  10 dB/floor for reinforced concrete  13 dB/floor for precast slab floor  26 dB isolation for corrugated steel (diffraction path dominates)

Chapter 4 –Mobile radio propagation( Large-scale path loss) 54 Dr. Sheng-Chou Lin

Wireless Communication

Acceptable Cellular Voice Quality

•AMPS VOICE QUALITY IS ACCEPTABLE IF OVER 90% OF THE COVERAGE: 1) VOICE S/N RATIO > 38 dB IN FADING ENVIRONMENT 2) RF CARRIER-TO-INTERFERENCE RATIO (CIR) > 18 dB •GIVEN 1) AND 2), 75% OF USERS GRADE THE SYSTEM AS “GOOD”OR “EXCELLENT” •MATHEMATICALLY:

S  C = 10 log 3 C + 15 dB, where > 13 dB, and > 0.6 N 10 4 I I S = BASEBAND SIGNAL-TO- RATIO N C = RF CARRIER-TO-INTERFERENCE RATIO I •IN DIGITAL SYSTEMS WITH VOICE COMPRESSION, VOICE QUALITY IS USUALLY QUANTIFIED PSYCHOACOUSTICALLY VIA Mean Opinion Score (MOS) RATINGS ON ASCALE OF 1 TO 5.

•AN MOS SCORE OF 3 IS CONSIDERED MINIMALLY ACCEPTABLE Bernardin, C.P. et al ,"Voice Quality Prediction in AMPS Cellular Systems using SAT," Wireless 94 Symposium, Calgary, July 12, 1994, pp 238-241.

Chapter 4 –Mobile radio propagation( Large-scale path loss) 55 Dr. Sheng-Chou Lin

Page 28 Wireless Communication

Lesson 4 Complete

Chapter 4 –Mobile radio propagation( Large-scale path loss) 56 Dr. Sheng-Chou Lin

Page 29